You’re not just imagining it: the Hillary Clinton vs Donald Trump vote totals do look rigged

As a political journalist, I hate empty conspiracy theories. I like to go where the bulk of the evidence is pointing. So even though I’m as shocked at Donald Trump’s victory over Hillary Clinton as anyone else, I have been unwilling to preemptively accuse the vote totals of having been rigged or altered. It’s taken me nine days of looking at the numbers and trends and patterns, but I’ve come to the conclusion that for once, the conspiracy theorists appear to have been right: this election looks rigged.

There is no single smoking gun in the voting results which led me to this conclusion of its own accord. Rather it’s the totality of all of what I’m about to lay out which has led me to the conclusion that there was indeed something fishy with the numbers, the sheer number of highly unlikely red flags. So bear with me as I lay out a mountain of evidence.

What first made me suspicious was the notably low general election voter turnout. Pundits have explained that away by pointing out that Clinton’s favorability rating was fairly low and Trump’s was even lower. But throughout the primary season, the American voters made clear that they very much wanted to turn out for or against these two candidates. And despite all the talk about favorability, these two candidates each got millions more primary votes than any of their challengers. So the notion that the 2016 election would receive two and a half million fewer votes than the 2012 election, which most Americans considered to be a far less consequential election, is odd. But the demographic breakdowns are what truly raise eyebrows.

Some have pointed to voter suppression in certain swing states as the supposed reason for lower voter turnout. But the 2012 election was subject to the same kinds of suppression. And for that matter, so were the 2016 primary contests. Despite the voter purging and long lines, we saw African-Americans in southern states come out of the woodwork to vote for Hillary Clinton in the 2016 democratic primary, even though they didn’t necessarily dislike Bernie Sanders. But according to the official tallies, those same African-Americans in North Carolina suddenly decided they no longer cared if Hillary Clinton won the 2016 general election, even though she was now facing an opponent in Donald Trump whom they viewed as an abhorrent racist.

Other states and other demographics raise other red flags. For instance, somewhere around seventy percent of all the votes cast in the 2016 general election in Florida were cast early. Various exit polls pointed to Hillary Clinton receiving as somewhere between 54% and 59% of the Florida early vote, much of it from Hispanic-Americans. That meant Trump would have had to have received somewhere between 59% and 71% of the election day voting in Florida in order to have caught up and tied the state. He would have needed between 62% and 74% of the election day vote to have won it by the one percent overall margin he supposedly won it by.

Although democrats tend to do somewhat better in early voting and republicans tend to do somewhat better on election day, it’s virtually impossible for Trump to have come back and won Florida after it was already basically in the bag for Clinton before election day even arrived. See my full mathematical breakdown for Florida voting in the middle of this article for why he simply couldn’t have come back to even tie up Florida, let alone win it.

There has also been much made of how the polls ended up being so wrong. Having spent the past year and a half observing the polls in this election, I fully agree that it’s easy for one or two polls to end up being very wrong. We saw it all the time in the primary season. But historically speaking, going back through the eighty years in which presidential polling has been conducted, it’s virtually impossible for the polling averages to have been this thoroughly wrong. In fact the last time they got a Presidential general election wrong outside the margin of error was in 1948 – and polling was unsophisticated crap back then.

But nevermind the overwhelmingly unlikely odds of the 2016 polling averages having been wrong. The more immediate trend is one which we saw during the primary season. In any given hotly contested primary state, Donald Trump tended to perform the same as, or worse than, his polling averages. We saw it in his very first contest in Iowa, where he shockingly lost despite being favored. We saw it again in Wisconsin and other states. In contrast, Hillary Clinton tended to perform about the same as, or better than, her polling averages in most states. For instance she was favored to win South Carolina by around twenty points and she won it by more than forty points.

In a general election matchup between one candidate who had spent the past year underperforming his poll numbers, and one candidate who had spent the year outperforming her polls numbers, the logically expected outcome is that Hillary Clinton would have won by the same amount she was ahead in the polling averages or more. Now she did win the popular vote in the general election by around one percent. But logically, based on the final polling averages and the existing pattern observed in the primaries, she should have won it by four points or more.

Finally, there were four swing states in which Hillary Clinton was definitively favored to win, but she ended up losing: Florida, Wisconsin, Pennsylvnia, and Michigan. In the final tallies she lost each of them by right around one percent of the vote. That’s not how numbers work. If Trump had won those four states legitimately due to pockets of voters that pollsters didn’t know about, we would have seen a more random dispersion of the results. Trump might have won one of those states by four percent, won another of them by two percent, lost another one percent, and so on.

It is statistically suspicious that in every state where Donald Trump pulled off an upset, he won it by right around one percent, just what he needed to win it, no more and no less. Results don’t naturally play out that way. The final tallies in Florida, Wisconsin, Pennsylvnia, and Michigan read like someone went through and nudged each of them just over the mark so Trump would win them, but didn’t want to arouse too much suspicion by giving him any larger of a victory in those states than he needed.

So where does all of the above get us? It’s a mountain of statistical and mathematical and logical and demographic discrepancy and suspicion and nothing more. I can’t definitively prove that the vote tallies were rigged. And as a practical matter it would be so tricky for a hacker to rig the results in various states, without any of the local precinct overseers catching on, that no one has even been able to posit a plausible method for pulling it off. But still, these things can’t all have legitimately happened.

In order to believe that the official vote tallies are legitimate, you have to accept that all of the above legitimately happened: African-Americans in the south went from turning out in droves for Hillary Clinton in the primary to not caring if she won the general election. Donald Trump got sixty-something percent of the same-day voting in Florida. The polling averages were wrong for the first time in modern history. Trump beat his poll numbers despite having spent the primary season tending to fall below them. Clinton fell below her poll numbers despite having spent the primary season tending to beat them. In every state where Trump pulled off a shocking upset victory, he just happened to do it with one percent of the vote. And in an election that everyone cared particularly deeply about, no one really turned out to vote at all. I can accept any one of the above things happening as an isolated fluke. I cannot accept all the above happening. And so for once in my evidence-driven career, I’m left to believe that the conspiracy theorists are right: the vote tallies are rigged.